Profit-driven Computation Offloading for Mobile Edge Computing in Wireless Metropolitan Area Networks

Abstract

The applications and services are growing computation-intensive vary with the advent of wireless communication technology, whereas the computing resources and capabilities of mobile devices (MDs) are finite. Mobile edge computing (MEC) is a compelling way to offload applications from MDs to the cloudlets for performing, by means of deploying cloudlets nearby the MDs, which is capable of alleviating the computing burden of MDs and reducing the completion time of applications. However, MDs are irregularly dispersed in the Wireless Metropolitan Area Networks (WMAN), and the service provider needs to recover the cost of the deployment and maintenance of cloudlets and achieve profitability. As a consequence, it poses a significant challenge to reasonably offload applications from MD to the proper cloudlet for MEC in the WMAN. In this paper, we study the user satisfaction of MDs, the profit of service provider and the resource utilization of cloudlets as a multi-objective optimization problem. In the sequel, we devise a profit-driven computation offloading method, under the collaboration of MD- cloudlet-cloud, to obtain the optimal offloading strategy such that maximizes the user satisfaction of MDs, the profit of service provider and the resource utilization of cloudlets. To this end, we conduct comprehensive experiments to validate the efficiency of the proposed method. The numerical results demonstrate the performance of the proposed algorithm is prominent.